loss function
A loss function quantifies the difference between predicted outputs and actual targets, guiding the optimization process by evaluating model performance during training.
- ChatbotID: Identifying Chatbots with Granger Causality Test
- ConfTuner: Training Large Language Models to Express Their Confidence Verbally
- Contrastive Self-Supervised Learning As Neural Manifold Packing
- DBLoss: Decomposition-based Loss Function for Time Series Forecasting
- Energy Loss Functions for Physical Systems
- Fully Dynamic Algorithms for Chamfer Distance
- Gatekeeper: Improving Model Cascades Through Confidence Tuning
- Improving the Euclidean Diffusion Generation of Manifold Data by Mitigating Score Function Singularity
- Learning to Instruct for Visual Instruction Tuning
- Solver-Free Decision-Focused Learning for Linear Optimization Problems